Every year, doctors around the world order roughly thirty million scans and treatments that depend on medical isotopes: faintly radioactive substances that either reveal disease under a scanner or carry a dose of radiation straight into a tumor. Almost no one outside of nuclear medicine knows how thin the supply really is.
Nearly the entire world supply of the most-used diagnostic isotope traces back to seven research reactors. Six of them were built at least fifty years ago, designed for Cold War science rather than for healthcare, and only later converted into global medical lifelines. They are now the weak point of the whole system, each one a mechanical fault or a budget cut away from going dark.
When one stops, hospitals feel it within days. The product cannot be stockpiled, because it decays within hours, so there is no reserve to fall back on. During the 2009 to 2010 shortage, world supply fell by as much as 70 percent. Scans were rationed and some patients were pushed toward surgery for no reason other than that the material was not there. It happened again in late 2024, when a Dutch reactor went offline for repairs while others were down for maintenance.
A policy problem, not a technical one
For decades, the producing reactors were government machines that sold isotopes below what they actually cost. On the surface this looked generous. In practice it kept private capital out: no company could justify building modern capacity while competing against a subsidized, state-backed reactor selling at a loss. So nobody built, and the world stayed tied to the old machines.
The alternative already exists, and it is running on the shores of the Great Lakes. Canada’s commercial CANDU power reactors can be refueled and loaded with isotope targets without ever shutting down, producing cancer medicine and electricity in the same shift. Bruce Power was the first operator anywhere to make a major therapy isotope inside a working power reactor. Ontario’s Darlington station became the first commercial plant licensed to produce the key diagnostic isotope.
Our dependence on a handful of aging research reactors was a choice. It can be un-chosen.
Why now
A new generation of cancer therapy is running straight into the supply problem. These treatments pair a scan isotope that finds the cancer with a strike isotope that kills it, and several are already approved in the United States and Europe. The most promising versions rely on actinium-225, which is so scarce that the entire world’s annual output can treat fewer than a hundred patients. In 2024, a late-stage clinical trial was paused for lack of it. The drug had not failed. The supply simply did not exist.
The case for smarter regulation
A little over a decade ago, most of the world’s diagnostic isotopes were made from weapons-grade uranium. A 2012 United States law set a clear goal, declined to dictate the method, and left industry to find the best technical route. Producers converted, officials certified sufficient non-weapons-grade supply by the end of 2021, and exports of the dangerous material were banned. Security and supply improved together, without costing patients a single dose.
That is the model. It applies to every bottleneck still in the way: licensing regimes that never imagined a power reactor making medicine, customs rules written for cargo that does not decay, and a payment structure that fails to reward reliability.
The briefing paper closes with a roadmap for policymakers covering regulatory pathways, pricing and reserve capacity, transport and customs, and a realistic assessment of what small modular reactors can and cannot yet deliver.
